Sebastien Loriot
Sebastien Loriot
Strange, I gave it a try in the CGAL lab demo and I could get all connected components. I'll have another look tomorrow directly with the code.
Indeed it does not work. @lrineau @janetournois I simply updated the CGAL example to add read from image: https://gist.github.com/sloriot/b4812ffda6c0c8d9780b0d6b8409258e Using the data shared by @CheLamVien [here](https://unibox.uni-rostock.de/getlink/fi8pdKv9C2hFX7ZKPF9Nk2/0720_GAG_Syn_2D_10ngml_IL-1B_J2_nucleus_clean.inr.gz) It does print that...
Successfully tested in CGAL-6.1-Ic-59
Successfully tested in CGAL-6.1-Ic-132
If we really want to get an option without PCA, we should also do a similar change for polylines (using squared length and one representative halfedge). What do you think...
> Yes, we can offer that as an alternative. Do you mind adding that to your TODO list? AFAIU Mael would need it.
Your issues here come from the fact that the boundary of your scan does not project without intersections with your surface, like here: One option could be to work on...
Thanks a lot for the report. It should be fixed by https://github.com/CGAL/cgal/pull/9142
Mesh_3 fully red in CGAL-6.2-Ic-60
@sven why work in progress?